Related Experiment Video
Updated: May 11, 2026
05:57
Long-term Video Tracking of Cohoused Aquatic Animals: A Case Study of the Daily Locomotor Activity of the Norway Lobster (Nephrops norvegicus)
Published on: April 8, 2019
Estimating detection and identification probabilities in maritime target acquisition
Jonathan M Nichols1, Kyle P Judd, Colin C Olson
1U. S. Naval Research Laboratory, Washington, DC 20375, USA. jonathan.nichols@nrl.navy.mil
Applied Optics
|May 15, 2013
Summary
This study presents Bayesian methods to estimate target detection and identification probabilities versus range. These approaches quantify uncertainty and efficiently model performance curves using experimental maritime data.
Area of Science:
- Detection and Identification Technologies
- Bayesian Estimation
- Probability Theory
Background:
- Estimating target detection and identification probabilities as a function of range is crucial for sensor performance evaluation.
- Quantifying uncertainty in these estimations is essential for reliable system assessment.
- Existing methods may not fully leverage available data or provide comprehensive uncertainty measures.
Purpose of the Study:
- To develop and demonstrate Bayesian approaches for estimating target detection and identification probabilities versus range.
- To analytically derive posterior probability distributions for performance parameters.
- To quantify estimation uncertainty using credible intervals and to efficiently model performance curves.
Main Methods:
- Adoption of a Bayesian estimation framework.
- Analytical derivation of posterior probability distributions for detection and identification probabilities.
- Development of credible intervals to quantify parameter uncertainty.
- Direct estimation of parameterized performance curves using Bayesian methods.
- Application to experimental data from wide field-of-view imagers.
Main Results:
- Credible intervals for detection and identification probabilities were derived and analyzed as a function of range.
- A second Bayesian approach efficiently estimated parameterized performance curves, yielding distributions of probability versus range.
- Both methods were successfully demonstrated using real-world maritime target data.
Conclusions:
- Bayesian estimation provides a robust framework for assessing sensor performance and quantifying uncertainty in target detection and identification probabilities.
- The developed methods offer efficient data utilization and provide valuable insights into performance variations with range.
- The study validates the proposed approaches using practical experimental data, highlighting their applicability in real-world scenarios.
